https://popups.uliege.be/3041-539x/index.php?id=1714 Index terms fr 0 https://popups.uliege.be/3041-539x/index.php?id=1948 A new approach for enlarging digital images is proposed. In existing approaches, the assumed reducing operators must be suitable ones for the methods, which means that desirable results are not obtained in other situations. Therefore, an enlargement scheme that can appropriately take reducing operators into account is needed. In this paper, we propose a new enlargement method that can be used for any reducing operators based on the framework of image restoration problems and estimation of the component that belongs to the kernel space of the reducing operator by using statistical properties of natural images. Wed, 17 Jul 2024 15:21:19 +0200 Mon, 07 Oct 2024 15:01:41 +0200 https://popups.uliege.be/3041-539x/index.php?id=1948 https://popups.uliege.be/3041-539x/index.php?id=1713 Image restoration is a typical ill-posed problem. Regularization technic represented by a regularized MP-Inverse filter (RMPIF) is widely used to deal with the ill-posedness. In order to derive the best performance of the filter, the parameter, which controls the regularizability, should be appropriately chosen. In this paper, we present a new criterion for parameter choosing based on modifying the subspace information criterion which is first proposed for model selection of supervised learning problems. Some numerical examples are also shown to verify the efficacy of the proposed criterion. Tue, 16 Jul 2024 12:03:55 +0200 Tue, 16 Jul 2024 12:04:06 +0200 https://popups.uliege.be/3041-539x/index.php?id=1713